Analysis Report — travel

Generated on 2025-10-06T23:10:50.122830
Rows: 8000
Positive: 34.3% • Negative: 40.6% • Neutral: 25.1

Executive summary

Overall sentiment: mean VADER compound = 0.08 (std 0.38).
Top positive terms: need, wonderful, convenient, class, paris, staff
Top negative terms: need, overbooked, paris, terrible, disappointing, non

Notable time points

Most positive period: 2025-08-20 (mean=0.108).
Most negative period: 2025-09-17 (mean=0.055).

Top Terms

Sentiment Over Time

Predictive terms (predicting negative sentiment)

The table below lists features (tokens) that are most predictive of the negative label based on a simple logistic model.
term coef
terrible 4.092615
overbooked 4.024035
disappointing 3.860367
lost 2.614753
non 2.495278
cancelled 2.269184
delayed 2.241576
refundable 1.368705
ticket 1.019783
oked 0.870235

Download CSV

Sentiment spikes

Detected spikes in rolling sentiment (z-score threshold). Review the rows to understand dates and severity.
timestamp sentiment_mean count rolling_mean rolling_std zscore spike
2025-07-09 0.075035 582 0.075035 0.000000 0.000000 False
2025-07-16 0.071834 616 0.073434 0.002263 -0.707106 False
2025-07-23 0.095979 627 0.080949 0.013114 1.146071 False
2025-07-30 0.078214 622 0.082009 0.012512 -0.303289 False
2025-08-06 0.092974 626 0.089056 0.009508 0.412125 False
2025-08-13 0.070676 614 0.080621 0.011343 -0.876852 False
2025-08-20 0.108087 622 0.090579 0.018821 0.930274 False
2025-08-27 0.085170 625 0.087978 0.018863 -0.148861 False
2025-09-03 0.081879 623 0.091712 0.014277 -0.688753 False
2025-09-10 0.085800 631 0.084283 0.002106 0.720507 False

Download CSV

Topic — sentiment correlation

Correlation between LDA topic weights and VADER sentiment (positive values indicate topics associated with more positive sentiment).
topic keywords corr_with_sentiment
topic_0 experience, airport, assistance, need, voucher, delayed -0.285531
topic_1 need, grand, hyatt, flight, refund, voucher 0.298171
topic_2 staff, helpful, process, reservation, destination, non 0.355382
topic_3 seat, check, scenic, terrible, disappointing, convenient -0.000130
topic_4 paris, want, confirm, booking, voucher, luggage -0.114817
topic_5 travel, airline, offered, hotel, luggage, refundable -0.199268

Download CSV

Clusters — top terms & samples

Clusters are produced by LSA embeddings + KMeans; top terms summarize each cluster and sample snippets give context.

Cluster top terms

Cluster 0: experience, dubai, sydney, rome, london, tokyo, york, new, cancelled, delayed

Cluster 1: assistance, airport, need, delayed, booked, cancelled, lost, rescheduled, enjoyed, confirmed

Cluster 2: want, confirm, paris, luggage, seat, voucher, hotel, destination, check, itinerary

Cluster 3: grand, hyatt, need, flight, hotel, refund, seat, reservation, booking, voucher

Cluster 4: airline, travel, offered, luggage, booking, hotel, destination, check, refund, itinerary

Cluster 5: staff, helpful, process, disappointing, terrible, refundable, non, overbooked, convenient, wonderful

Cluster samples

Cluster 0:
hotel paris lost experience scenic
flight new york upgraded experience first-class
hotel tokyo booked experience non-refundable

Cluster 1:
flight rescheduled airport need assistance
refund enjoyed airport need assistance
refund upgraded umm airport need assistance

Cluster 2:
want know confirm luggage terrible luggage paris
want confirm itinerary disappointing itinerary paris ticket:827967
want confirm hotel first-class hotel paris

Cluster 3:
wonderful check-in grand hyatt need get check-in
non-refundable flight grand hyatt need get flight
wonderful voucher grand hyatt need get voucher

Cluster 4:
airline upgraded itinerary offered convenient travel itinerary
airline cancelled luggage offered overbooked travel luggage ticket:452203
airline delayed seat offered know terrible travel seat

Cluster 5:
hotel process convenient staff helpful
seat process non-refundable staff helpful
seat process terrible staff helpful

Download cluster top termsDownload cluster samples